Welcome everybody to the SiTime Fireside Chat. My name is Mark Lipacis. I'm the senior semiconductor analyst at Evercore ISI. We're very excited to have Beth Howe, who's the Executive Vice President and CFO of SiTime. She joined in November of 2023. We'll set this up as a fireside chat with a Q&A. If we have time at the end, maybe we can take some questions. I would say, Beth, the biggest question we get is what is your exposure to AI, and how fast is that actually growing? Could you give us a little bit of a outline of how big is that AI business and how you see it progressing?
Well, first, thanks for having us today.
Okay.
It's good to be here. AI is definitely one of those exciting trends that we are taking advantage of. As you know, we created the category of precision timing to really unlock and disrupt this 70, 80-year-old market called timing. It's an $11 billion TAM. Clearly one of the most exciting parts of it is the AI and the data center. Some of the products that we developed for it, a couple of years ago, for communications and telecom, as well as for the data center, have really been taking off as we've seen the whole explosion in data centers. As we look at it, one of the things that's exciting about SiTime in general is the diversification of our applications, both in AI as well as beyond. If I think about the AI opportunities, they really are very broad.
You can think about them generally around two dimensions of compute or processing, as well as networking or the connectivity fabric. Timing is required in both, whether it's the compute trays, XPU, TPU, GPU, or across the networking topologies, whether that's cables, active electrical cables, opticals, switches, NIC cards, and I'm sure we'll dig into. Really the diverse range of applications across AI require timing and increasingly high-performance precision timing.
Got you. You have a reporting segment called CED, Communications, Enterprise, and Data Center, that's about 2/3 of your business. Is there a way to get a little more resolution on that or to just to break out like the data center portion? Is that possible to do?
Sure. As we think about it, and in Q1, you're right, it was about 2/3 of the business. Again, Q1 had some interesting dynamics. If we look at all of 2025, it was 53% of the business, and it's growing faster than our other categories, so I expect it to grow in terms of mix. I would think about it more in the 50%-60% range on an annualized basis. As we look into that, we do include communications, enterprise, and data center. Data center or AI is the vast majority of that revenue.
As I was talking about earlier, within that, there are really a range of applications. Transceivers and optical cables is a piece, but also associated with compute, switching, NICs, are all applications that contribute to that revenue and that strong growth that we've seen, I think, now eight quarters of triple-digit growth.
Right. Okay. Think about it in the kind of the 50% range for the total group and data center being the vast majority of that.
Okay, great. Then on the call, I think you also mentioned that you're expecting strong shipments for your oscillators at 400 Gb and 800 Gb for the next two years. That sounds like a lot of confidence and conviction, and I'm wondering, could you help us understand where is that coming from? Are your customers giving you Is this a forecast that you get from them or is it hard orders? How do you get that level of conviction?
I think we have been working closely with our end customers to get more and more visibility and their forecast four quarters, in some cases, eight quarters out of where they're seeing the growth and the opportunities. That gives us confidence as they are giving us more confidence based on their forecast. The other piece of it is, I think you mentioned 400 and 800 Gb and 1.6 Tb. We are leading in 1.6 Tb as that's rolling out. In some ways, I think, more fast quickly than folks had initially expected and really ramping here in 2026. We do expect that 400 and 800 will continue to also be very strong growers for this year and next year.
Mm-hmm. How do we think about the pricing for the higher-speed solutions? You also talked on your call about the Elite 2 Super-
TCXO
TCXO, right. How do we think about your pricing dynamics as you go generation to generation and improving the functionality?
We do see ASP increases as we move from gen to gen and the higher performance. Typically, the higher bandwidths require higher frequencies, those higher performance characteristics do garner higher ASPs. It can vary from application to application, you can think about probably in the 20%, 30%, maybe even 40%. It'll depend on the specific application, we do see that performance and those features generating higher ASPs.
When you double the speed, it doesn't necessarily double the ASP, something like 20%-40% or something like that.
Probably more in that range.
Right.
Yeah.
Got you. How do you manage the supply chain? When you talk about, I think this last quarter, 88% year-on-year growth, right? You can't turn around without hearing about some kind of a supply tightness or capacity issue. Can you talk about your main suppliers? I mean, your fabless, right? How do you manage the risk of tightness or not getting product?
We are fabless and we are outsourced, so we have very good relationships with our wafer suppliers as well as our OSATs. I think the top-line message is we've been working very closely with them, and they're supporting us from a forecast perspective and getting us what we need. That's work that our operations team does every day in terms of working with our suppliers, whether it's TSMC for CMOS wafers or Bosch for our MEMS wafers or our OSATs. We've had long-working relationships, and we continue to work very closely with them as we've seen our forecasts increase and to be able to get the supply and to generate the forecasts and give them the visibility so that we can support our customers as their forecasts have increased.
Your customers or your customer customers are signing up for long-term agreements with their suppliers and if you look at an NVIDIA or an AMD, their purchase commitments have gone way up also. Is this something that's happening for you? Have you gotten to the point in the supply chain where you're prepaying, or you have to make hard commitments that creates a liability on your balance sheet?
It varies from vendor to vendor, but we are working very closely with them, and so in some cases, it may be we're putting in POs for longer periods in order to secure the wafer supply that we need. In some cases, we may talk with them about sharing CapEx in order to build out lines to get equipment, in order to build out the lines that we need more from a back-end perspective.
I think those are good economic trade-offs in terms of making sure that we can then support our customers. I think the other thing to keep in mind is that the geometries are 55 nm, 180 nm kinds of geometries, so we're not in the sub-10 nm kinds of geometries, which I think helps. Because of the nature of the programmability of our products, we do tend to kind of keep more wafer stock and as raw materials and then can then translate that into finished goods based on specific customer demands.
Having your own inventory, building up your own inventory is part of that risk management process.
That's not new for us. We always, especially given the size of our company and the size of our customers relative to us, we have always had a strategy to maintain healthy inventory wafer stocks in order to be able to supply those customers and give them the confidence that we can meet their demands for whatever their product categories are.
Mm-hmm. What has happened with your lead times?
Our lead times are fairly consistent. We haven't seen a significant change in our lead times. Again, I think back to our programmability, the quality of our products, our supply chain, we're able to be very resilient and turn around products, in some cases, on even shorter lead times when customers require it.
Got you. I think another highlight in your quarter was the aerospace and defense market, and I believe satellite. Satcom is a big part of that. Can you tell us what you're seeing there? What is the opportunity?
Within aero and defense, again, there's a diversity of applications, and it's another exciting area for lots of reasons. We are in a number of different low Earth orbit satellite providers, and so have customers across the different providers of satellites. The content there is really interesting. It can be up to $2,000 of content per satellite. As you think over the next couple of years, thousands, maybe up to 10,000 satellites being launched into orbit, a really nice opportunity for us. It's also more broad than that.
If we think about defense and the kind of increasing need for technology for whether it's navy or boots on the ground in the army, our products, especially our Epoch OCXO, allows for a characteristic called holdover, which means when you're having GPS spoofing or GNSS spoofing, which is increasingly issue for military and other applications, the fact that they can maintain that holdover for up to 24 hours allows them to know where their troops are, to coordinate on the ground, even in the absence of having GPS connectivity, which is a big benefit for many militaries around the world.
Got you. $2,000 per satellite is your TAM? That's how to think about it, or that's what you designed right now?
Design wins up to $2,000. It varies a bit from customer to customer, but you can think about that kind of order of magnitude in terms of the SiTime timing content in those satellites.
We talk about seven to 10,000 satellites that launch over the next three years. Can you give us a sense of your market share?
I don't have a good sense of our market share. I think in general, we're less about market share, more about how do we provide customers value, how do our high-performance products really solve problems for them and really enable the capabilities that they're looking for in their products.
Got you. Okay. Can we shift over to the consumer market? If you think about what's going on downstream and maybe one of your customers is more vertically integrating, how does that impact you? On the consumer side, if you talk about customers doing their own vertical integration and on different parts, not necessarily the timing parts.
Consumer has been a vertical for us for some time. We've got one large consumer customer in particular, but also other customers where they see value in precision timing for the solutions, whether it's smartwatches or other applications for consumer. One of our large consumer customer, we did have a design win associated with their modem, we're shipping two chips associated with that modem. As that proliferates into their product portfolio, we would expect significant growth there as they roll that out.
It's a share gain effectively for you.
Yes, but I'd frame it a little differently. They're making a transition as they've done with other parts of their stack from using third-party technology to developing their own technology.
Right.
Really it's about that transition from a third-party technology to their internally developed technology, which we're attached to.
Right. You were not attached to the third party?
Correct.
Okay. That sounds like market share gain to me. Okay. The other exciting space I think a lot of people have been talking about is physical AI. How big is this for you today? What's the opportunity?
Physical AI is, I think, an emerging opportunity. I think we're still trying to get our arms around exactly how big it can be, but we see opportunities, and depending on how broadly you define physical AI, clearly humanoid robots is evolving, right? Amazon's talking about having over a million robots in their warehouses. You've got Optimus, you've got a number of different providers that are developing humanoid robots, and those require precision timing, in terms of the sensors and the actuators and the synchronization that's required. There's an opportunity for us in those kinds of robotics. As well as more broadly, you think about autonomous driving and physical AI coming up in lots of different areas, whether it's in manufacturing, whether it's in warehousing, whether it's in some of these other autonomous kinds of applications.
Got you. We're just starting in this-
Early innings. Yeah.
Right. Got you. Can you also talk about the consolidation in the industry, the Renesas Timing acquisition? How does that expand your scale and your portfolio?
We're really excited about the Renesas Timing Division acquisition. It really is a transformational milestone for the company in terms of our scale. As we've said for many years, we wanted to expand our portfolio. If you think about timing, there's three product categories, resonators, oscillators, and clocks, and clearly most of our revenue comes from oscillators. We do have a clocking business as well that's been growing. We really wanted to get scale in clocking so that we had a full portfolio for a wide range of customers. We have talked for many years about wanting to do an acquisition in this space and the Renesas business that comes out of IDT and was ICS before that, is really a premier brand and the largest timing provider. It gives us scale in a new way. We have about 50 timing clocking products today.
They have over 500 and so we're really excited about having that full breadth of portfolio. It's also exciting because of the complementary nature of the markets and applications. About 75% of their business is in what we would call comms enterprise data center, they call it cloud and comms, and the other 25% is in automotive and industrial. They're exposed to some very attractive markets. From a financial perspective, it is also an accretive business for us. They generate north of 70% gross margins, accretive operating margins as well. We're really excited about the opportunity to invest in this business and close it, and bring it into the fold.
Excellent. We are running out of time and I was hoping, you talk to a lot of investors, the stock has been doing very well, so I always like to ask, what do you think is the biggest disconnect between your perception of your fundamentals and where the investment community is? What is the market missing about SiTime?
I think we're, in general, underappreciated. I think some of the things that are underappreciated about SiTime are the growth opportunities ahead of us, both organically as SiTime, and then once we close the Renesas' timing business acquisition, the ability to integrate that and really accelerate our path to a billion dollars. I think a year ago, when we said, "Hey, we want to be a billion-dollar or multibillion-dollar company," people were maybe a little skeptical. I think they're a little less skeptical today, I think the best quarters and years are still ahead of us in terms of the opportunities to grow in our timing, not only in AI and data center, but also across these broad range of applications, whether it's autonomous driving and automotive or consumer or industrial or aero and defense. Just the range of opportunities.
Financially, we've got a lot of operating leverage in the model, and so our ability to continue to grow at the strong growth rates and then deliver operating leverage and cash flow to the bottom line are also really strong.
The intent for the cash flow that flows out of the bottom of the model.
Yeah
How do you think about that?
We really think about investing in the business, and that's really our highest priority, is continuing to grow. I think one of the things that we're doing right now is really investing in go to market as well as R&D, so that we have feet on the street to go after these opportunities, right? To really capitalize on all of the growth that's in front of us, and then reinvesting back in R&D so that we have the product portfolio, not only today, but in the next three to five years to support that growth of the next generation of capabilities.
Excellent. I think we're at time here. Beth, thanks for joining us today and sitting down and having a chat and sharing your insights about SiTime.
Thanks, Mark. It's good to be here.
Thank you.